Become pro at transistor theory and practice

Become pro at transistor theory and practice
Posted by matrixofdynamism on 28 Dec, 2023 06:31
When we learn about transistors we first start with basics of semiconductors and leard about the diode and p-n junction. Then we move over to transistors, BJT and FET.

My question is, what are the specific topics/subjects names that must be fully comprehended and mastered and their order, for a person to be considered pro at BJT and FET design. Being pro means that we can analyze circuits containing these components and design new circuits that use them. Being pro here does not mean understanding all details of the solid state physics as a physicist but rather understanding the transistor circuit design art as an engineer.

#1 Reply
Posted by ppTRN on 28 Dec, 2023 09:27
You should take a look at "the art of electronics". It basically explain everithing with planty of examples

#2 Reply
Posted by G0HZU on 28 Dec, 2023 12:03
I've got the second edition of the Art of Electronics (I've not read the third edition) and whilst it's a brilliant book, it doesn't really do much more than analyse transistors at DC and audio frequencies.
The input impedance will change a lot up at RF, and sometimes the input can show a negative resistance up at RF and this is bad for stability.

The stability of transistor circuits doesn't get much (if any) mention at all. This is one area where I think pretty much all textbooks are weak on. Even the RF theory books are weak on stability analysis.

Many people won't be aware of the significance of this, because stability doesn't really get analysed in much depth in text books. If it does get analysed, it typically uses basic narrowband techniques at the frequency of interest.

A few simple VNA tests will show that the little jellybean 2N3904 BJT can go unstable up to about 1GHz when biased up at about 10V and 10mA for example. The MPSH10 BJT can go unstable to about 2.4GHz. The classic 2N3819 or BF256B JFETs can go unstable up to about 900MHz or so. So a stability analysis for these transistors really should be carried out well into UHF if a designer wishes to be thorough.

If a 2N3904 BJT circuit was oscillating at (say) 800MHz, then the person might not realise this if their test equipment doesn't cover up to 800MHz.

Instability is obviously also possible at lower frequencies. To understand 'why' these devices can go unstable at various frequencies in a particular circuit is a key skill to master. There's obviously lots of other aspects of transistor theory and design that need to be mastered, but instability is often overlooked.

#3 Reply
Posted by watchmaker on 28 Dec, 2023 15:08
I disagree with the recommendations for "Art of Electronics" as a source for LEARNING. It was one of the first books I bought and while I think it is an important reference, it is not for learning from scratch.

You sound as if you want to pursue a path similar to the one I am pursuing. Basically self study for a two year tech level or even a full EE course.

The upside of self study is that there is no pressure to perform. The downside is you do not have classmates from which to learn. But this group goes a long way if you are not too needy.

You can see my recs in the sticky on "Primers"; it is one of the last. To that list I need to add: Boylestad's Electronic Devices and Circuit Theory. I needed every one of these sources to understand DC ckt analysis and finally not panic when I have a ckt to solve. This skill is important for future understanding.

I wound up creating one notebook with explanations of how to use each of the analyses.

I will soon start the L and C parts. So far it has taken 3 months. Bear in mind I am retired and I can spend as much time as I want.
I try to spend at least 10 hours a week. You DO need to brush up on algebra and matrix algebra. I am waiting till next summer to learn calculus. For these Brilliant and Kahn Academy are excellent.

To keep myself motivated, I have also read up on basic semiconductor theory and transistor circuits. Watch Dave's youtube videos which are linked at the very bottom of this blog page. He provides very clear instruction. In fact, his is the only place I saw a drawing of a FET with the insulator between the channel and gate which explains why a FET is voltage controlled. Awesome.

Now that I understand determinants (Matrix alg for solving linear equations), I acquired a TI 89. I am not interested in becoming a math major, I just want to understand what is going on behind the scenes so I can apply the tools. I also will be using WulframAlpha and MatLab as tools for calculations.

Be careful. Many learning sources use very simple ckts to demonstrate the various components of analysis. This can lead to a false sense of mastery. This is where the Boylestad comes in.

Spoiler alert. Bolyestad provides some very efficient techniques. Maybe it is because it was the path I followed, but I really do suggest leaving Boylestad as a capstone. For me, it all suddenly came together.

You will want /need to learn a ckt simulator like MicroCap, pSpice for TI or MultiSim. All of these can be had at no cost if you look around. Each has its own quirks although MicroCap seems to have the least (from my POV).

You will need the simulator to build the practice problem and check your result after you have done the problem. It is too hard to find complete solution sets for the problems in these books.

At some point you will want to build physical circuits. Bread boarding. Probably the least expensive way is with Analog Discovery which is a PC based device that has scopes, meters, voltage sources, signal generators etc that interact with your breadboard for real.

This is what is used in many educational programs. I do not think you can break it!

I do think the Diligent you tube series "REal Analog" is a good starting point. But, you then absolutely need to branch out to the other sources. I have even used the Heathkit Education systems books to good effect.

Probably the least helpful is the USN NEETS course (through which I was taught tube theory 50 years ago). The modified course leaves out "hard" topics and the course really relies on classroom interactions.

I have gotten much help and encouragement on this forum, so if I can offer any advice on a learning path, I am very willing to share.

Maybe more than you want, but there it is.

Regards,

Dewey

#4 Reply
Posted by Jwillis on 28 Dec, 2023 21:08
To understand transistors, you need to know the three classifications, types of each classification, operation modes of each type and classification, characteristics and application.
The three classifications are BJT ((bipolar junction transistors), FET (Field Effect Transistors), and IGBT(Insulated-Gate Bipolar transistors), each having their own types,sub types, operation modes and applications.
You can search the internet for each kind of transistor to get the extensive information much easier than someone could explain here condensed in one post.
And you can always visit here when you have questions. Many knowledgeable individuals here would be happy to help out. No such thing as a stupid question.
Good luck and have fun.

#5 Reply
Posted by the_cake_is_a_lie on 30 Dec, 2023 16:28
Quote from: watchmaker on 28 Dec, 2023 15:08
I disagree with the recommendations for "Art of Electronics" as a source for LEARNING. It was one of the first books I bought and while I think it is an important reference, it is not for learning from scratch.

For real, that book is not for beginners. Good for me when I learned transistors in a classroom setting and can get hazy on things. I do agree on the concept of learning with a textbook. Internet tutorials make assumptions on what you already know, what you should know and what you have a 15-minute attention span for to like / com / sub and Patreons get SPICE models! They're a complement.

Being a "pro" is somewhat subjective. The Electronics II course for juniors in EE expects students to be able to hand solve any 2-transistor circuit in BJT or MOSFET or JFET form. Electronics I is where 1-transistor common X circuits and use as amps and switches, including logic gates, is taught. At some point the concept of CMOS gets thrown in. Intro to Computer Engineering teaches building computer memory from flip-flops and latches, including with a clock on a breadboard. With all that under your belt, you should know when to use a BJT versus FET and look at a 3-stage amp circuit diagram and have a decent idea of what each stage is doing.

Anyone who can do all that is a pro to me. Time to use Level 2 formulas with exponents, logs and Vt with temperature in Kelvin. I think much of electronics is having strong fundamentals. Maybe I'm a pro even though I've never touched IGBTs or expensive GaN or SiC MOSFETs. I don't like high power electronics or motors. You don't have to know it all. We tend to specialize in niches.

Annoying to me to see internet people use BJTs for everything who spread suboptimal designs on GitHub cause Beta math is easy and 90% of the transistor tutorials are for them. If you can use MOSFETs at all, amp something besides audio copy pasta, don't make beginner mistakes and debounce switches then you're in the top 10% of hobbyists. Maybe that's pro level. Oh parasitic capacitance is limiting the bandwidth? Time for cascode to defeat the Miller Effect.

Sorry for writing so much and for complaining. This topic gets personal.

#6 Reply
Posted by Terry Bites on 30 Dec, 2023 21:14
I think you need to take two parallel courses here.
One in semiconductor physics and another in practical circuits.
Knowledge of the former will not necessarily lead to success in the latter.
In fact it can just be a distraction.
Build up your practical experience to gain confidence then dive into the nitty gritty.

I found it very telling at uni, there were hobyists who'd decided to go pro and ee virgins who thought that a dregee would make it all clear.
Guess who's projects flopped!

#7 Reply
Posted by ommsiva on 04 Jan, 2024 15:59
You can take " semiconductor approximation by malvino " or electronic principles by the same author.

Good introduction of concept with less math.

I am eager to get list of books to become professional.

#8 Reply
Posted by bostonman on 04 Jan, 2024 16:37
If I can make some recommendations and assuming you're willing to spend a few dollars.

Get yourself a transistor kit. I don't know the cost, but I'm sure you can find a kit that has an assortment of basic transistors for under $15. Also, a resistor kit that has some basic values (you can series/parallel them to get different values).

Then look into a cheap adjustable power supply and multimeter.

Now you can measure the exact resistor values, note them, set up various basic circuits, and take measurements. You'll know the resistor values from measuring them (update - I meant you'll know the exact values because you measured them, rather than going by the marking and introducing the tolerance into the calculations), the voltage, then you can get the current. At this point you'll know the transistor gain, etc...

If you blow the transistor, you have more in the kit.

#9 Reply
Posted by bostonman on 04 Jan, 2024 19:02
eBay has a 600 piece transistor kit for $10.61. Also it looks like they have several (cheaper) multimeters for under $30.

They also have resistor kits for under $10. I imagine you can also find a 30v, 1 or 2A adjustable power supply for cheap too.

For maybe $50 to $75 you can set up circuits, take measurements, read about them, and gain hands on experience. Personally I learn better if I tinker with real stuff and then read rather than read first and try to visualize what I'm learning.

#10 Reply
Posted by mawyatt on 04 Jan, 2024 20:04
The transistor in various flavors are the most complex single devices man has ever created, and likely will ever create, thus requiring an enormous amount of time and effort to understand at the rudimentary level much less master as a Pro.

If one wishes to become a "Pro" with understanding Transistors, then consider taking undergrad courses at a university. These should include fundamental Semiconductor Theory and basic Electronic Circuits wrt the transistor (BJT, JFET, MOS). Once these courses are mastered, then move onto graduate school and take Advanced Semiconductor Theory with Active Device emphasis, and/or Advanced Transistor Devices/Circuitry Theory and Practice which should include detailed analysis and design of active devices like BJT, MOS, JFET, IGBT and so on. Then you can study the various Semiconductor Technologies such as Ge, Si, SiGe, GaAs, SiC, InP, GaN and how they are utilized with the various active device types mentioned, as well as supporting circuitry for these drives types and semiconductor technologies.

This gives one a more detailed overall "look" into how these active devices operate and behave, then you begin to realized the inherent complexity of such and how these devices are actually fabricated for use. Realizing you have just scratched the surface and the True Learning Experience begins, you can truly appreciate the enormous efforts by others in this field.

At this point one realizes the entire process starts all over again with Integrated Circuits, and another career long learning curve which is never ending is awaiting!

Best,